Hi Gareth, Jianhua, Silvan and others, Thanks for all your comments and suggestions. I have previously made = SU-8 structures using combinations of Omnicoat, SU-8 2002, 2005 as seed = layer and then built thicker SU-8 structures on them up to a total = thickness of 150um without any delamination. My dehydration process = were as Gareth suggested. =20 However the above were all done on SiO2 wafers. I have not tried these = process on glass though. Will try and let you know. =20 Jeffrey ________________________________ From: mems-talk-bounces@memsnet.org on behalf of Jianhua Tong Sent: Tue 10/16/2007 4:07 AM To: gjenkins@f2s.com; General MEMS discussion Subject: Re: [mems-talk] SU-8 on glass if you just want to make a mold to fabricate some PDMS structure, you = can make a seed layer use SU8-5, or SU8-10. Spin SU8-5 to form a 3-4 microns layer, bake, and expose to UV(without = mask), postbake, and develop, then bake it in oven at 175C for 2 hours. Then = you can spin and pattern other thick SU8 layer without adhesion problem. Quoting Gareth Jenkins: > Omnicoat is certainly a good idea but good adhesion is also possible > with the following steps: > > Treat with conc. H2SO4 (I find this as good as Piranha and safer / > more eco-friendly since you don't need to make it fresh each time). > Thoroughly dehydrate. 200-220C for 30mins on a hotplate. > Ensure even and complete crosslinking with a higher exposure dose than > recommended. > Take account of stress during the PEB. If large areas are being > exposed, consider stress relief structures in the design. A lower > temperature PEB for longer time can greatly reduce delamination > effects.